How to implement clustering to improve data prediction ?

Hi all, i had built a predictive model and through the filtering of data, removal of unimportant data variables as well as removal of outliers, my predictive accuracy is sort of minimized to what i expected. But i heard clustering can be used to improve predictive accuracy or reduce the impact of outliers. I would like to know how can i implement clustering to my predictive model ?

Answer 1

[Image Analyst]

Browse through the demos, examples, documentation, and webinars here: http://www.mathworks.com/products/statistics/?s_tid=hp_fp_list